The Myth of the Rational Policy User

What's Perplexing:

There is a common myth that much policy and service design still assumes that citizens would read instructions carefully, weigh options calmly, and complete multi-step processes if it benefits them.

The reality is more complicated. Behavioural economics and public administration research suggest that stress, scarcity, cognitive overload, and low trust are some reasons that change how people decide and whether they follow through. When policy journeys are complex, the cost is often paid in attention, time, and error. These costs fall hardest on people with the least bandwidth to spare. In effect, governments can end up designing for a citizen who does not exist in reality.

I believe this topic matters for public servants not because people are irrational, but because administrative complexity becomes a hidden gatekeeper. It shapes who completes a process, who drops out, who makes mistakes, and who disengages entirely.

Why this myth persists even when we know better?

The rational-citizen assumption is resilient because it is organisationally convenient. Rules are often written for legal clarity, auditability, and risk control. Service steps are added to prevent edge cases. Responsibilities are split across teams, agencies, and contractors. Each individual addition can look reasonable, yet the combined experience can become unmanageable for the people the policy is meant to serve.

There is also a scaling problem. Behavioural interventions often show promising results in pilots where context is controlled and staff attention is high. When scaled, they encounter variation in local capacity, IT systems, languages, user needs, and frontline discretion. In essence, the science may be sound, but institutions may lack the delivery conditions required to reproduce results consistently.

What the evidence says: scarcity and burden are design constraints:

Two research traditions are especially relevant to understand this case.

First, behavioural economics shows that people under pressure do not simply try harder. Scarcity of money, time, or stability tends to narrow our attention to urgent demands and reduces our capacity for planning, form-filling, and error-checking. A well-known experimental study found that financial strain can measurably reduce cognitive performance in the short term (Mani et al., 2013). This is indeed a design reality. If a process assumes surplus attention, it will systematically exclude those living with the least of it.

Second, public administration scholars describe this exclusion through administrative burden. Burden includes the learning costs of figuring out what to do, the compliance costs of doing it such as forms, documents, and appointments, and the psychological costs such as stigma, fear of penalties, and distrust (Herd and Moynihan, 2014). When burden rises, take-up tends to fall, even when the benefit is valuable for individuals.

A concrete illustration comes from higher education finance. In a large randomized evaluation, simplifying completion of a complex aid application increased successful submissions and downstream enrollment (Bettinger et al., 2012). The key lever was not a new incentive or a new message but it was reducing friction so people were most likely to not stall.

These patterns also appear in everyday government work across sectors. Means-tested benefits often require periodic re-certification. Even short lapses can trigger churn where eligible recipients lose access temporarily because paperwork is too complex and time-consuming. Public health programs can see steep drop-off when booking, eligibility rules, and reminders are fragmented across channels. Housing and homelessness services frequently require repeated assessments across agencies, turning access into a test of persistence rather than need. These issues are indeed common to whenever complex processes meet constrained lives of individuals.

An alternative: Cognitive Load Audits:

If the myth is that better incentives and clearer guidance will do the work, a practical alternative is to design for low bandwidth and high variability programs. A useful tool here is a cognitive load audit, a short, structured review of how mentally demanding a policy journey is for a real person on a difficult day.

A cognitive load audit does not require a new unit or a major transformation program. It can be embedded into routine policy and service design with a modest time investment.

The audit asks basic questions such as how many steps are required, how many decisions must a person make, how many documents are needed and how hard are they to obtain, how much reading and interpretation is required, how often must someone switch channels or repeat information, and where are the points of no return where small errors trigger delays, penalties, or loss of access.

Identifying the cognitive load turns a vague issue such as the process was confusing into something design teams can measure, compare, and reduce. It also helps identify where simplification will produce the largest equity gains, especially at steps where time pressure, language barriers, disability, unstable housing, or limited digital access make completion least likely.

Governments have legitimate obligations around integrity, safety, and stewardship of public funds. The question is whether controls are targeted where they reduce risk most, or spread broadly in ways that primarily screen out eligible people. In many contexts, targeted verification, clearer thresholds, and fewer repeated proofs are both more humane and more effective.

Thus, a cognitive load audit will not solve problems driven by supply constraints such as a shortage of housing units or clinic capacity, but it can prevent scarce resources from being wasted through avoidable drop-off and rework. It is most helpful where the main barrier is navigation, follow-through, and repeated compliance.

Keynote:

The key idea is that governments often design as if citizens have unlimited time, attention, and trust. Evidence suggests the opposite is frequently true, especially for the people most affected by policy. Embedding cognitive load audits into policy and service design is a low-cost way to make systems work for real human decision-making, not an idealised version of it.